We propose to study the extent to which the cell cycle and pharmacokinetic properties of normal rapidly proliferating tissues can be used to minimize tissue and whole animal toxicity during multiple dose treatment with cell-cycle-phase-specific (CCPS) anticancer agents like cytosine arabinoside (ara-c). In the best of all worlds, both normal tissue and tumor properties would be utilized to maximize the therapeutic advantage in cancer treatment schedules. Unfortunately, it is becoming increasingly clear that heterogeneity within each tumor and between tumors will limit the effectiveness of generalized strategies based on tumor properties. Thus we propose to concentrate instead on the normal tissues. Understanding the principles behind normal tissue toxicity should allow rational anticancer strategy development in experimental systems and facilitate the extrapolation of positive findings to the clinic. To accomplish this we will determine the correlations between crypt cell depletion and animal survival, and between pluripotent stem cell depletion (CFU-S) and animal survival. We will measure for both tissues, the durations of DNA synthesis inhibition, extent of stem cell kill, the time dependent levels of ara-c and ara-ctp, and the crypt and CFU-S cell cycle durations. We will use these data in a mathematical model to caluclate the dose level at which to administer ara-c. We will also predict the extent of crypt cell and CFU-S cell damage as a function of dose interval time in a ten dose treatment course. We will then experimentally measure crypt cell and CFU-S damage and the median animal survival time as a function of dose interval time following ten doses of ara-c administered at the calculated level. Since the predicted dose interval for low host toxicity will differ from that actually observed, we will investigate changes in cell cycle kinetic and pharmacokinetic properties of the crypt cells and CFU-S cells which may occur during the course of a ten dose, moderately toxic ara-c treatment schedule. Our existing model will then be corrected on the basis of the measurements. The degree to which the corrected model calculations agree with the experimentally measured tissue damage and animal survival times vs dose interval will provide a definitive measure of the extent to which cytokinetic and pharmacokinetic data can be used in the design of cancer therapy schedules for CCPS drugs.